Do Australian sclerophyll forests exhibit seasonality? an analysis with phenocam, eddy covariance fluxes, and satellite derived phenology.

Abstract:

Temperate broadleaf evergreen forests in Australia exhibit characteristically unique and contradictory seasonality behaviors with strongly seasonal gross primary productivity (GPP) values and weak to no seasonality in satellite-derived vegetation indices (VIs), leaf area index (LAI), and fraction of absorbed photosynthetically active radiation (fPAR).

As part of adaptation strategies to highly variable rainfall and water balance deficit conditions, sclerophyll forests allocate carbon to long-lived, thick leaves with low water content, and generally exhibit small seasonal changes in canopy infrastructure (LAI). Erectophile leaf angle distributions, and/or differences in leaf adaxial and abaxial optical properties allow the leaves to achieve thermal protection. However, these leaf traits complicate any spectral analysis and the study of sclerophyll forest phenology.

Our goal was to utilize tower mounted phenocam imagery of whole-canopy, multiple tree crowns, and understory layers to trace multi-functional phenology profiles at three sclerophyll forest sites (one banksia dominated and two eucalyptus dominated) all part of the Terrestrial Ecosystem Research Network (TERN).

We contrast and compare in-situ phenocam time series data with satellite vegetation products from the Moderate Resolution Imaging Spectroradiometer (MODIS), and eddy covariance measures of ecosystem built photosynthetic capacity (ecosystem light use efficiency, LUE, and chlorophyll fPAR).

We found that at sclerophyll forests, despite ecosystem photosynthetic capacity exhibiting weak seasonality, climate and in particular rainfall pulses, drove diverse responses over each of the different forest components (e.g. overstory and understory). Interestingly, tree and understory growing and browning cycles were out-of phase, and contributed to the characteristic VI seasonality behavior of the whole ecosystem.